In a carefully worded announcement on Monday, Apple said John Giannandrea, who has been the company’s AI chief since 2018, is “stepping down” and will exit after an advisory period through spring.
A New Leader With Google and Microsoft Experience
His replacement is Amar Subramanya, a Microsoft executive who previously spent 16 years at Google, most recently leading engineering for the Gemini Assistant. The hire brings deep familiarity with Apple’s biggest competitors.
A Shake-Up Fueled by Apple Intelligence Missteps
The leadership change follows a turbulent debut for Apple Intelligence, which launched in October 2024 to mixed reviews. A notification-summary tool produced inaccurate headlines in late 2024 and early 2025, prompting complaints from publishers including the BBC.
Siri’s much-publicized overhaul also faltered. A Bloomberg investigation reported that Apple’s software chief Craig Federighi found key features not working just weeks before launch, leading to an indefinite delay and class-action lawsuits from iPhone 16 buyers.
Internal Tensions and Leadership Reassignments
Bloomberg reported that Giannandrea had already been sidelined by early 2025. Siri was transferred to Vision Pro creator Mike Rockwell, and Apple’s robotics division was removed from his oversight. The investigation cited weak communication between departments, budget issues, and morale problems that led some staff to nickname the group “AI/MLess.”
The report also documented a steady loss of AI researchers to rivals such as OpenAI, Google, and Meta.
Leaning on Google to Fill Capability Gaps
Apple is now expected to rely on Google’s Gemini to power the next version of Siri — a notable shift given the companies’ long-running rivalry across mobile, cloud, and AI.
Giannandrea joined Apple from Google, where he ran Machine Intelligence and Search. At Apple, he directed AI strategy, machine learning infrastructure, and Siri’s development.
Subramanya’s Task: Help Apple Catch Up
Subramanya will now take over those responsibilities and report to Federighi with the goal of accelerating Apple’s AI capabilities.
A Distinct Approach With Trade-Offs
Apple continues to pursue a privacy-first AI strategy built around on-device processing using Apple Silicon. More complex requests are routed through Private Cloud Compute, which processes and then immediately deletes user data.
The trade-off is that Apple’s on-device models are smaller and less capable than the large models running in competitors’ data centers. Apple’s limited collection of user data has also pushed it toward licensed and synthetic datasets rather than the vast real-world corpora used by its rivals.
Whether this philosophy will ultimately differentiate Apple or leave the company behind remains an open question.
Featured image credits: Getty Images for Techcrunch
For more stories like it, click the +Follow button at the top of this page to follow us.